Multiple observational studies demonstrate bariatric surgical approaches to obesity lead to substantial and sustained weight loss and have dramatic impact on type 2 diabetes mellitus (T2D), including remission of hyperglycemia in 77% and substantial reductions in diabetes medications in 84% of cases. Whether bariatric surgery represents a first-line therapeutic strategy for T2D management, particularly in those with lesser degrees of obesity has become a major question in the medical field and the subject of three recent international consensus conferences. Likewise, the optimal surgical procedure to propose for treatment of less obese T2D patients remains controversial. Minimal level 1 data is available to guide these approaches, with only one randomized trial demonstrating greater glycemic improvement and weight loss one year following LAGB compared to medical (non-surgical) management. Yet metabolic improvements are greater following Roux-en-Y gastric bypass (RYGB) then LAGB and disproportionate to weight loss suggesting weight independent mechanisms. Furthermore recent data suggest LAGB may have complications leading to high removal rates over time. Concurrently, significant advances in pharmacologic and dietary approaches and a greater understanding of optimal medical and lifestyle management programs to treat T2D raises the question of why surgical options should be considered as primary therapy at the very time that medical, non-surgical treatment has improved efficacy. In this context we are conducting an ongoing prospective randomized NIH sponsored trial [DK086918] to compare either LAGB or LRYGB to an intensive multidisciplinary medical and weight management program (IMWM) on rates of normalization of dysglycemia at one year (diabetes resolution defined as fasting plasma glucose levels <126 mg/dL and HbA1c <6.5%). We have a unique and time-sensitive opportunity to evaluate and compare durability of surgical and intensive medical intervention strategies, and to simultaneously and comprehensively evaluate hormone and metabolite factors that promote metabolic improvement in our randomized controlled prospective cohort undergoing clinical phenotype evaluation. In this application we propose to extend observations out to 3 years, a time when weight stability has been achieved. We will test the hypothesis that Roux-en-Y gastric bypass will be more effective to improve glycemia, weight and cardiovascular risk biomarkers at the 3 year time point than intensive medical management, but that laparoscopic adjustable gastric band will not. Our primary endpoint will be normalization of dysglycemia at 3 years. Our study will provide data essential to inform clinical decision making. Furthermore, we hypothesize changes in circulating enteroendocrine hormones, bile acids, and/or branched chain amino acids contribute importantly to the magnitude and durability of improved glycemic control. We will employ detailed informatics to identify patterns predictive of clinical response to address these important mechanistic questions.